46-2
Cisco Catalyst Blade Switch 3130 for Dell Software Configuration Guide
OL-13270-01
Chapter46 Configuring Fallback Bridging
Understanding Fallback Bridging
A VLAN bridge domain is represented with switch virtual interfaces (SVIs). A set of SVIs and routed
ports (which do not have any VLANs associated with them) can b e c onfigured ( gro uped toge the r) to
form a bridge group. Recall that an SVI represents a VLAN of swi tch ports as one interf ace to t he routing
or bridging function in the system. You associate only one SVI with a VLAN, and you configure an SVI
for a VLAN only when you want to route between VLAN s, to fallb ac k-b ridg e non rout ab le pro toco ls
between VLANs, or to provide IP host connectivity to the switch. A routed port is a physical port that
acts like a port on a router, but it is not connected to a router. A routed port is not associated with a
particular VLAN, does not support VLAN subinterfaces, but behaves like a normal routed port. For more
information about SVIs and routed ports, see Chapter 10, “Configuring Interface Characteristics.”
A bridge group is an internal organization of network interface s on a sw it ch. You cannot use bridge
groups to identify traffic switched within the bridge group outside the switch on which they are defined.
Bridge groups on the switch function as distinct bridges; that is, bridg ed traffic and bridge protocol data
units (BPDUs) are not exchanged between different bridge groups on a switch.
Fallback bridging does not allow the spanning trees from the VLANs being bridged to collapse. Each
VLAN has its own spanning-tree instance and a separate spanning tree, called the VLAN-bridge
spanning tree, which runs on top of the bridge group to prevent loops.
The switch creates a VLAN-bridge spanning-tree instance when a bridge group is created. The switch
runs the bridge group and treats the SVIs and routed ports in the bridge group as its spanning-tree ports.
These are the reasons for placing network interfaces into a bridge group:
To bridge all nonrouted traffic among the network interfaces making up the bridge group. If the
packet destination address is in the bridge table, the packet is forwarded on a single interface in the
bridge group. If the packet destination address is not in the br idge table, the packet is flooded on all
forwarding interfaces in the bridge group. A source MAC address is learned on a bridge group only
when the address is learned on a VLAN (the reverse is not true). Any address that is learned on a
stack member is learned by all switches in the stack.
To participate in the spanning-tree algorithm by receiving, and in some cases sending, BPDU s o n
the LANs to which they are attached. A separate spanning-tree process runs for each configured
bridge group. Each bridge group participates in a separate spa nning -tre e i nsta nce. A brid ge group
establishes a spanning-tree instance based on the BPDUs it receives on only its member interfaces.
If the bridge STP BPDU is received on a port whose VLAN does not belong to a bridge gr oup , th e
BPDU is flooded on all the forwarding ports of the VLAN.
Figure 46-1 shows a fallback bridging network example. The switch has two ports configured as SVIs
with different assigned IP addresses and attached to two different VLANs. A nother port is configured as
a routed port with its own IP address. If all three of these ports are assigned to the same bridge group,
non-IP protocol frames can be forwarded among the end stations connected to the switch even though
they are on different networks and in dif ferent VL ANs. IP addresse s do n ot need to be ass igned to routed
ports or SVIs for fallback bridging to work.